Re: [xml-dev] Stick with XML ... JSON is a minefield of securityrisks an

On Wed, Nov 2, 2016 at 5:45 PM, Jim Melton <jim.melton@oracle.com> wrote:

Roger,

Thanks for this message and for the link to your paper. I have a few remarks to offer, if I may.

JSON is not inherently "bad", nor is it inherently "good". Neither is XML, nor COBOL, nor SQL (my particular standardization specialty), nor ice cream.

Mirrors precisely the current state of political dialogue in the USA. My candidate is flawed but I can point to flaws in your candidate too. The effect of that type of argument is to sweep the scale and nature of the respective flaws under the carpet so that everybody can be placed on a false equal footing.

Too many people seem to grab hold of the latest notion and believe that it might finally be the "silver bullet" that solves all of our problems, simply and elegantly.

Brings to mind another topical political analogy. It starts with B ends in t and has an X in the middle.

Eventually, the hype wears off, intellectually honest people see the reality, which is that all of those things are merely tools.

But just as with social political and economic realities, Eventually is often too late and it is more a case of act in haste repent forever.

JSON has an important place in the information technology ecosystem, which is acknowledged by the support that both XQuery and SQL have given it.

The motivation for such support often doesn't extend to more than alot of people want it now and we have to provide it to make sure we stay relevant and/or don't lose market share.

Nobody would question that say SQL has an important place in the IT ecosystem, but there are huge question marks over peoples ability to discern the scope and nature of what that place is. Often times it reduces to I know X and have applied it successfully over there, therefore I am going to conclude that it is a good idea to apply it over here.

Never mind that over there simple primitive data types where fine but over here we have been using a much larger and richer type system.

Never mind that over there no schemas were fine because that was because the schema-ish stuff was being taken care of over here.

Never mind that over there were didn't have to talk to same scope and variety of people that we need to talk to over here.

Let's continue to indulge the historical inability of technologists to acknowledge that the applicability of tools/solutions has boundaries and the continued refusal to assess the applicability of a tool/solution to a totally different domain.

The silver bullet thing is trite and hackneyed.

Look this thing is a piece of cardboard". Cardboard is and will remain very useful, but if you paint it a shiny color and go around telling people that it a suitable foundation for something I am supposed to sit on, instead of hiding behind no silver bullets I am going to call you out for it.